Perplexity rejects Reddit’s lawsuit alleging large-scale scraping of comments; company says it lawfully accesses public threads and doesn’t train foundation models. Key details and industry context.
Perplexity has pushed back strongly after Reddit filed a federal lawsuit accusing the AI sector of “industrial scale” scraping of Reddit comments for commercial model training. The complaint, lodged in New York on October 22, names Perplexity and three data-services firms, alleging that millions of user comments were extracted without permission to build commercial products. Perplexity says the allegation misunderstands its role and business model: it describes itself as an application-layer service that summarises public discussions and links back to original Reddit threads rather than training foundation models directly.
In its public rebuttal the company framed Reddit’s suit as part of a larger negotiation push — saying the action appears timed to influence licensing talks between Reddit, Google and OpenAI. Perplexity emphasised that it lawfully accesses publicly available content and that it cannot sign licensing agreements for model training because it does not perform that function. The firm also accused Reddit of trying to strong-arm smaller players into paying for access and vowed not to be used as leverage against competitors.
The lawsuit targets not just a chatbot maker but the web-scale infrastructure used by the AI industry — proxy services, scraping vendors and API providers — underscoring a growing legal scrutiny of data-acquisition practices. Reddit has previously licensed its data to major model makers, creating a new revenue stream; the platform says unauthorised scraping undermines that market. Industry watchers note this case could set precedents about public-web data, permissible indexing, and the responsibilities of downstream applications that surface user-generated content. Relevant search phrases to capture readers: “Perplexity Reddit lawsuit explained,” “did Perplexity steal Reddit data,” “AI scraping legal cases 2025,” “Reddit licensing AI models,” and short-tail queries like “Perplexity lawsuit” and “Reddit data scraping.”
